Jan 6, 2025: Our weekly Institutional Insights explores the latest market developments — market performance, industry news, exchange-traded fund (ETF) flows, trending topics, upcoming events and token unlocks — to help you supercharge your crypto trading.

Weekly Highlight — Markets Snap Five-Day Loss Streak 

Investors are now seizing the opportunity to buy the dip in U.S. stocks after a five-day decline that erased over $1 trillion in market value. Wall Street has broken its longest losing streak since April, with the Nasdaq-100 rebounding 1.7% and the S&P 500 rising 1.3% on Friday, January 3. These gains slightly offset the week's sell-off, which had continued into the new year following a tough December.

Despite concerns about slowing earnings growth among tech giants like the Magnificent Seven, traders remain optimistic, especially regarding AI leader NVIDIA Corporation. The market believes that, as long as retail investors keep investing in AI, the associated stock market boom is expected to persist. 

Stocks reached their session highs in the afternoon after Mike Johnson was reelected as speaker of the U.S. House of Representatives, indicating potential support for a business-friendly agenda. Meanwhile, bond yields rose, with the 10-year hitting 4.6% following comments from Richmond Fed President Tom Barkin about maintaining restrictive rates. U.S. manufacturing showed modest growth, with new orders at their highest since early last year, while investors assessed economic strength against the possibility of slower interest-rate cuts.

Weekly Crypto Highlight — AI Agent Spring Coming

Many investors are eagerly anticipating an Alt Season, a period when altcoins typically surge in value. However, current market dynamics suggest we may not see a widespread altcoin rally. The sheer number of tokens, combined with limited liquidity, has led to what's known as altcoin dispersion, making it crucial to select the right sectors and tokens.

, which are thriving even as the broader market declines. This nascent sector is experiencing rapid advancements, and its associated tokens face less pressure from venture capital unlocks. Many of its top protocols lack centralized exchange listings, indicating more buyers may enter the market.

Those investors who missed previous opportunities, such as the DeFi summer of 2020 and the Alt Layer 1 trade, now have another chance. Instead of targeting individual AI agents, the best risk/reward strategy lies in investing in infrastructure plays, which serve as the foundational tools for the sector.

The leading AI agent player is Virtuals Protocol, a platform for launching tokenized AI agents on the Base network. By purchasing tokens like Luna by Virtuals (LUNA) and VaderAI (VADER), users gain governance rights over specific agents, which allows them to both influence development and share in profits. 

Recently, daily protocol fee revenues dropped significantly in mid-December 2024 as activity slowed for the holidays. However, speculation around agent-associated tokens surged as the new year approached. Currently, Virtuals Protocol generates over $1 million in daily fees, though this is 50% less than figures for early December. Despite these challenges, the VIRTUAL token has more than doubled in value, reaching a market capitalization of $4.6 billion.

Other highlighted opportunities include Zerebro, which is pivoting to a full-stack framework for AI agents; ARC, with its Rust-based framework; and 

 (GRIFFAIN), a natural language on-chain search tool. These projects are well-positioned for future growth in the evolving landscape of AI technology.

 as we delve into leading AI agent tokens and the past week’s market movements. 

Advanced

Bybit Institutional_Weekly Poster.png

Weekly Institutional Insights: Market Snaps Five-Day Loss Streak; AI Agent Spring Coming

Our weekly crypto derivatives analytics report delves into macro events, the current state of crypto and trading signals from spot trading volume, and futures, options and perpetual contracts.

December's end-of-year options expiration brought an unexpected drop as realized volatility fell to the lowest levels of the month. Perpetual swap open interest remained stable, indicating these markets weren't heavily used for hedging against expired options, likely contributing to the muted volatility. 

While most expired open interest in Bitcoin (BTC) and Ethereum (ETH) hasn't been reinvested, the balance of calls and puts remains consistent. Notably, ETH options are pricing in higher implied volatility as compared to BTC, particularly for shorter expirations, leading to a flat term structure for ETH, while BTC is offering cheaper options exposure.

In the perpetual markets, low volatility and minimal spot movements have led funding rates for nearly all tokens to mirror broader market sentiment throughout the holiday season. On the options side, BTC's term structure is noticeably lower and steeper than ETH's, with realized volatility now dipping below implied volatility levels, highlighting an intriguing shift in market dynamics.

Please check out the report’s highlights.

Open Interest Solid Amid Year-End Options Expiration

Although open interest for both BTC and ETH perpetual swaps hasn’t returned to early December 2024 highs, it remained robust throughout the significant end-of-year options expiration event. This suggests that these markets weren’t heavily utilized to 

 of the expiring options contracts, likely contributing to the lack of volatility observed during this period. Instead, trading volumes have dipped noticeably during the winter holiday season, coinciding with a collapse in realized volatility back to December 2024’s lowest levels.

BTC Option Curve Remains Steep During Call-Put Parity

The expiration of December's end-of-year options didn’t trigger the anticipated surge in volatility; instead, realized volatility has plummeted to the lower end of its recent range. The implied volatility term structure remains steep compared to both its recent behavior and ETH's, with longer-dated implied volatility around 57%, and 1-week at-the-money options trading about five points lower. Much of the expired open interest hasn’t yet been rolled into new positions, maintaining a neutral balance between puts and calls. Consequently, BTC's options market positioning is relatively neutral, lacking the leverage it had at the start of December 2024.

Huge ETH Option Expiring Doesn’t Cause Volatility

A late December spike in realized volatility didn’t carry into the new year, with ETH’s spot price currently exhibiting volatility below that of short-tenor implied volatility. Interestingly, the implied volatility term structure has undergone a steepening and re-flattening over the past week, contrasting with BTC’s slightly steeper profile. This suggests that ETH’s options market is anticipating higher short-term volatility in spot price movements. 

Despite the significant expiration of options open interest at the end of December 2024, which didn’t lead to a rollover into later expirations, call options continue to dominate the market and have increased at the start of 2025.

block-scholes-x-bybit-report.jpeg

Bybit X Block Scholes Crypto Derivatives Analytics Report (Jan 2, 2025): BTC Options Continue to See Call-Put Parity, With Lower Short-Term Volatility Expected

Dive into FLock, a decentralized platform that revolutionizes the way AI models are developed, trained and deployed.

Artificial intelligence (AI) is poised to transform countless industries, but its development often raises questions about privacy, control and inclusivity. 

 is a unique decentralized platform that democratizes AI training, fine-tuning and inferencing. With its blockchain-powered infrastructure, FLock ensures transparency, security and community-driven decision-making, paving the way for AI models that align with established societal values. 

In this article, we’ll explore how FLock decentralizes AI training, its key features and its potential impact on developers, communities and users.

 platform that revolutionizes the way AI models are developed, trained and deployed. By leveraging blockchain technology and a novel approach called 

, FLock addresses critical issues in AI development, such as centralized control, lack of transparency and privacy concerns.

At its core, FLock’s mission is to democratize AI by enabling communities to participate directly in the creation and refinement of AI models. Through federated learning, FLock trains AI models across multiple decentralized devices or servers. This method ensures that data remains local, eliminating the need for raw data exchange while still enabling robust model training. By decentralizing the training process, FLock not only protects individual privacy, but also fosters a more inclusive ecosystem in which diverse contributors can collaborate.

What sets FLock apart is its integration of 

. This ensures that all contributions — whether data, compute power or feedback — are transparently recorded and fairly rewarded. Blockchain tech’s immutable ledger enhances trust among participants and safeguards the process against manipulation. Additionally, the platform’s on-chain incentive mechanisms encourage participation from a wide range of contributors, ensuring that rewards are distributed equitably.

Ultimately, FLock empowers communities to develop bespoke AI models tailored to their specific needs. By decentralizing the process and prioritizing transparency and collaboration, FLock breaks the monopoly of large corporations, creating an AI ecosystem that aligns with public ethics and societal values.

, an Oxford alumnus with a robust background in AI and blockchain technology. Sun, who serves as FLock's CEO, has an impressive career trajectory, having worked as the Director of AI for the Royal Bank of Canada and as an AI Research Fellow at Imperial College London. His vision for FLock stems from a deep understanding of both the technical and ethical challenges of AI development.

The founding team also includes other Oxford alumni, boasting over a decade of combined experience in AI research and the cryptocurrency industry. This team brings a collective blend of academic excellence and practical expertise that positions FLock to lead the decentralized AI movement.

FLock has successfully raised $11 million in funding across two rounds, reflecting strong investor confidence in its vision. In May 2024, FLock secured a $6 million seed round co-led by Lightspeed Faction and Tagus Capital, with additional backing from Digital Currency Group (DCG), Volt Capital, OKX Ventures and Inception Capital.

Later, in December 2024, FLock announced a $3 million successful funding round led by DCG and joined by prominent names like Animoca Brands, Fenbushi Capital, GnosisVC and GSR. These investments are being channelled into enhancing the platform’s testnet and federated learning client, accelerating FLock’s mission of enabling secure, collaborative AI development.

With this funding and over 1,500 nodes contributing from platforms like Numerai and Kaggle, FLock is rapidly scaling its impact on decentralized AI training.

FLock AI revolutionizes AI development through its decentralized training framework. This approach allows participants to contribute to AI model training without ever sharing their personal data. Instead, models are trained locally across devices using federated learning, which ensures that data remains with its owners while still enhancing the accuracy of the model overall. 

FLock’s reward system is designed to incentivize active participation. Contributors who provide data, computational power and/or feedback are compensated with FLOCK tokens, ensuring a fair distribution of rewards. 

FLock empowers its community through a robust governance framework. Decisions regarding AI model development, deployment and ethical considerations are driven by the community, ensuring alignment with established societal values. Through blockchain-based voting mechanisms, contributors can influence the direction of AI projects, fostering transparency and public accountability. 

FLock’s modular architecture allows contributors to participate in specific aspects of AI model development. Whether providing raw data or computational power or even refining algorithms, the platform enables a collaborative and scalable development process. 

At the core of FLock’s philosophy is its commitment to privacy and security. By leveraging blockchain technology, the platform ensures that no single entity controls the AI development process. Data remains local and protected by cryptographic measures, while the immutability of blockchain guarantees transparency in every transaction and model update. 

 begins with a blockchain and an AI layer.

The blockchain layer underpins FLock’s incentivization and security mechanisms.

The AI layer facilitates model training, refinement and deployment, leveraging both centralized and federated approaches.

Let’s have a look at the supply and demand for the FLOCK token.

FLock adopts a 1:2 ratio model between team/investors (33.3%) and community (66.7%), prioritizing community-driven growth and ecosystem expansion. The majority of tokens are allocated to incentivize user participation and long-term ecosystem success.

FLock aims to improve the future of AI development by decentralizing the training process and ensuring transparency, privacy and inclusivity. By leveraging blockchain and federated learning, FLock empowers communities to contribute to AI creation without compromising on privacy. Its innovative reward system and community-driven governance foster a fair and collaborative ecosystem. 

With significant backing and a growing network of contributors, FLock is positioning itself as a leader in decentralized AI, offering a more ethical and secure alternative to traditional AI development models. As it continues to scale, FLock is set to transform the way AI models are built and deployed.

 — and earn a share of a 10 million FLOCK prize pool. For more details, visit 

Intermediate

flock.png

FLock AI: Decentralized AI Training

In this tutorial, we’ll explore how you can complete a series of tasks in Network3 for Bybit Web3 Airdrop Arcade.

Task 1: Register for a Network3 Account and Connect Your Bybit Wallet 

: Navigate to the campaign page and click on the 

task. Click on the task, and Galxe will ask to redirect you to the Network3 register page.

: Register for a Network3 account using your email address. You may indicate a 

Task 2: Download the Network3 App and Run a Node for One Hour

 and choose your operating system to download the Network3 app.

 or powerbutton, depending upon which operating system you’re using. For ease of reference, we’ve used macOS and iOS as examples below.

: Keep running the node until the timer reaches 

Beginner

network3 tutorial.png

How to Complete Network3 Tasks for Bybit Web3 Airdrop Arcade

Trading

Explainers

Want to know if your crypto trading strategy works? Backtesting is a key component to assess the effectiveness. Learn how it works here.

 trading can be a daunting task. Crypto backtest is a powerful tool that allows traders to optimize their trading strategies and navigate this unpredictable landscape with confidence. Are you ready to unlock the power of crypto backtesting and elevate your trading game like a pro? Let’s dive in!

Backtesting serves as a critical process in which historical data is used to test a trading strategy, in order to evaluate its potential and fine-tune it for future application. It helps traders uncover the potential of their strategies and make necessary adjustments before investing real capital.

With the right approach, backtesting can be an indispensable tool in a trader’s arsenal.

The Role of Backtesting in Crypto Trading

 performance heavily depends on backtesting historical data. This process holds significant importance in the world of crypto trading, as it enables traders to recognize the strengths, weaknesses, and potential hazards of their strategies before investing real capital. By analyzing historical price data, traders can gain valuable insights into how their strategy would have performed in different market conditions and adjust their approach accordingly.

The backtesting process assists traders in recognizing potential strengths and weaknesses, refining their systems, and ultimately building confidence in their decision-making processes. However, backtesting relies on certain factors and is not without its limitations and challenges, such as the need for a significant amount of data to be accurate and the potential influence of overfitting and bias in the results. Using diverse data sets and out-of-sample data for validation can help mitigate these issues.

Although backtesting offers various benefits, it does have certain limitations, such as the unpredictability of future market conditions. While backtesting provides valuable insights, it does not guarantee future performance due to changing market dynamics and additional factors in live trading such as slippage, 

, and execution delays. Scenario analysis can be used to test the strategy under different market conditions and improve its robustness.

Another challenge is the potential for overfitting or bias in the backtesting results. Overfitting occurs when a trading strategy is excessively optimized to match historical data, yet fails to demonstrate effective results in actual trading scenarios. Bias, on the other hand, refers to the inclination to modify the model until an ‘ideal’ result is obtained, which in turn leads to unrealistic expectations and unsatisfactory performance in actual trading. To avoid overfitting and bias, traders should incorporate diverse data sets, utilize out-of-sample testing, and refrain from excessive parameter optimization.

Manual and automated are the two main approaches for crypto backtesting. They offer distinct advantages for testing a cryptocurrency trading strategy. Each method comes with its advantages and drawbacks, and the choice between them will depend on a trader’s individual requirements and circumstances.

Analyzing past price data is a requisite for manual backtesting. This helps evaluate the trading strategy to gauge how it would have fared in the past. This process is similar to paper trading, where traders simulate their trades without risking real capital. Manual backtesting can be time-consuming and prone to errors, but it also offers the advantage of a deeper understanding of the strategy and its underlying mechanics.

To facilitate manual backtesting, traders can use spreadsheets to compute various statistics such as:

Manual backtesting can help traders evaluate the performance of a backtesting strategy using historical data, recognize any flaws or weaknesses, and instill confidence prior to implementing backtesting trading strategies in real-time trading.

Contrarily, automated backtesting leverages technology and coding languages to assess trading strategies with increased effectiveness and precision. This method is particularly useful for backtesting complex strategies that require a large amount of data and computational power. Some popular platforms for automated crypto backtesting include MetaTrader 4/5, FXOpen’s TickTrader, and Cryptohopper.

The primary difficulty associated with automated backtesting is having the technical aptitude to code the strategy or having the resources to employ someone who can. However, automated backtesting is capable of rapidly and precisely processing a large amount of data, thereby reducing the possibility of human error. Python, C++, and Java are commonly employed in automated crypto backtesting.

Price, volume, and order book data are crucial for crypto backtesting. These types of data provide valuable insights into market trends and liquidity, helping traders make informed decisions about their strategies.

Price data sources include historical price charts and spreadsheets, which can be obtained from exchanges or third-party providers. The quality of price data plays a significant role in the accuracy of backtesting results, and it is crucial to obtain data from reliable sources and verify its accuracy.

Some sources that can provide accurate and reliable price data for crypto backtesting include Tradewell and Cryptodatadownload. In addition to obtaining data from exchanges, traders can also use APIs provided by platforms such as CryptoCompare or CoinGecko, which offer historical and real-time data at various time intervals.

Volume data is crucial for understanding market depth and liquidity. It refers to the total number of cryptocurrency units bought and sold within a specific time period on a particular exchange or across multiple exchanges. Monitoring volume data can help traders assess market trends, identify potential buying or selling opportunities, and make informed trading decisions.

High trading volumes can suggest increased interest and participation in the market, while low trading volumes may indicate a lack of market activity or limited liquidity. By analyzing volume data, traders can ascertain the level of liquidity in the market and recognize areas of support and resistance.

Order book data offers a more accurate representation of market conditions, allowing traders to simulate bid-ask spreads, slippage, and liquidity when testing their strategies. An order book is a compilation of snapshots of the market, including the price, volume, and depth of buy and sell orders.

Obtaining order book data for backtesting can be challenging, as the data may be difficult to procure due to storage costs for exchanges. A developer must collect and store data from the exchange itself, or they can access the order book snapshots through a third-party service. As a result, they have to ensure they have their data source available.

Utilizing order book data in backtesting provides a more precise demonstration of market conditions, permitting traders to make well-informed decisions regarding their strategies.

Selecting the Right Tools for Crypto Backtesting

The selection of appropriate tools for crypto backtesting hinges on various factors, including personal preferences, technical abilities, and the intricacy of the trading strategy. In this section, we’ll explore popular backtesting platforms and criteria for selecting the most suitable tool for your needs.

Popular crypto backtesting platforms include Altrady, Holderlab, Gekko, and Zenbot, each offering unique features and capabilities. Altrady is a professional trading platform specifically designed for cryptocurrency trading, offering smart trading, real-time market data and alerts, connectivity to multiple exchanges, and scaled ladders for an optimized trading experience.

Holderlab’s backtesting platform offers features such as portfolio optimization, correlation analysis, a trading bot for rebalancing, and an intuitive and advanced interface. Gekko is an open-source trading bot that enables users to implement and evaluate their trading strategies automatically, while Zenbot is a command-line cryptocurrency trading bot compatible with multiple digital assets.

By exploring the features and capabilities of these platforms, traders can determine the most suitable backtesting tool for their needs.

When selecting a backtesting tool, factors such as ease of use, customization options, data quality, and community support should be considered. A user-friendly backtesting tool should offer:

Data quality is essential for obtaining precise backtesting results, and traders should ensure that their backtesting tool offers reliable data from reputable sources. Community support can also play a significant role in the efficacy of a backtesting tool, providing an avenue for:

By considering these factors, traders can select the most suitable backtesting tool for their needs.

Analyzing and Interpreting Backtesting Results

Following the completion of the backtesting process, the subsequent step involves analyzing and interpreting the results through forward performance testing. Evaluating key performance metrics can help traders identify strengths and weaknesses in their trading strategy, allowing them to make necessary adjustments for improvement.

Key performance metrics, such as the success (win) ratio, Sharpe ratio, maximum drawdown, and equity curve, can help traders assess the effectiveness of their strategies. The Sharpe ratio is a metric that measures the performance of an investment strategy relative to a risk-free rate. It’s calculated by taking the excess return of a strategy and dividing it by its standard deviation. A higher 

 indicates that the strategy is more effective at compensating for the risk taken on the investment.

The Sortino ratio, another important metric, measures the risk-adjusted performance of an investment or trading strategy, focusing on downside volatility and taking into account only the standard deviation of negative returns. A higher Sortino ratio suggests that the strategy is more effective in avoiding significant losses and achieving the desired return.

By evaluating these key performance metrics, traders can optimize their trading strategies, minimize potential risks, and manage trading costs effectively.

Identifying strengths and weaknesses in a forex trading strategy allows traders to optimize their approach and minimize potential risks. By analyzing backtesting results, traders can pinpoint areas where their strategy excels and areas where it falls short. This information can be used to make necessary adjustments to the strategy, ultimately improving its performance and increasing the likelihood of success in real-world trading scenarios.

Continuous improvement and optimization involve refining the crypto trading strategy based on backtesting results and staying up-to-date with market trends and conditions. By regularly iterating and refining their strategies, traders can ensure that their approach remains relevant and effective in the ever-changing crypto market.

For successful crypto backtesting, it’s important for traders to evade overfitting and bias, ensure data quality, and continuously improve and optimize their trading strategy. By following these tips, traders can ensure that their backtesting efforts yield valuable insights and help them make more informed decisions in the crypto market.

Overfitting and bias are common pitfalls in backtesting that can lead to inaccurate results and unreliable trading strategies. Overfitting occurs when a strategy is excessively optimized to match historical data, yet fails to demonstrate effective results in actual trading scenarios. To avoid overfitting, traders should use diverse data sets and out-of-sample testing, which can help validate the strategy’s effectiveness in a broader range of market conditions.

Bias, on the other hand, refers to the inclination to modify the model until an ‘ideal’ result is obtained, which in turn leads to unrealistic expectations and unsatisfactory performance in actual trading. To avoid bias, traders should be cautious of excessive parameter optimization and ensure that their trading strategy is based on sound principles rather than fitting the data too precisely.

Ensuring data quality is crucial for accurate backtesting results. Poor-quality data can lead to erroneous and misleading analysis, ultimately resulting in ineffective trading strategies. To ensure data quality, traders should obtain data from reliable sources, such as reputable exchanges or third-party providers, and verify its accuracy.

In addition to obtaining accurate price data, traders should also consider volume and order book data, which can provide valuable insights into market depth and liquidity. By using high-quality data and verifying its accuracy, traders can ensure that their backtesting results are reliable and meaningful.

Continuous improvement and optimization involve refining the trading strategy based on backtesting results and staying up-to-date with market trends and conditions. By analyzing and interpreting backtesting results, traders can identify the strengths and weaknesses of their trading system and make necessary adjustments for improvement.

Furthermore, remaining aware of current market trends and conditions helps traders adjust their strategies to better suit the dynamic crypto market. By consistently adapting and refining their trading strategies based on new data and market conditions, traders can enhance their backtesting performance and ultimately, their success in the crypto market.

Crypto backtesting is an essential tool for optimizing trading strategies and navigating the volatile world of cryptocurrency trading. By understanding the backtesting process, selecting the right tools and data sources, and analyzing and interpreting results, traders can fine-tune their strategies and minimize potential risks. With continuous improvement and optimization, traders can confidently tackle the ever-changing crypto market and maximize their trading potential.

 is a great option for backtesting cryptocurrency as it allows users to perform backtests without writing code and supports backtesting against 5, 20 and 50 years of historical data.

 which contains an extensive collection of crypto futures historical data enabling users to backtest and optimize trading strategies.

What is the most profitable strategy in crypto?

 is the most profitable strategy in crypto. It involves exploiting the price differences between various crypto exchanges and platforms to make money. Furthermore, the lack of regulation on the crypto market allows traders to take advantage of triangular arbitrage and other forms of arbitrage trading.

What is the main purpose of crypto backtesting?

The main purpose of crypto backtesting is to evaluate a chosen trading strategy using historical data in order to optimize it for future use and minimize potential risks.

What are the two primary methods for crypto backtesting?

The two primary methods for crypto backtesting are manual and automated backtesting. Both provide valuable insight into how a given trading system may perform under different market conditions.

Unlock the Power of Crypto Backtest: Master Your Trading Strategy like a Pro

Grab Up to 5,000 USDT in Rewards

Grab Up to 5,000 USDT in Rewards